ACKR3 agonism induces heterodimerization with chemokine receptor CXCR4 and attenuates platelet function

Abstract

Background: Platelet receptors ACKR3 and CXCR4 play a crucial role in a variety of cardiovascular diseases. Like most chemokine receptors, CXCR4 is a G protein coupled receptor that induces platelet activation. In contrast, the atypical chemokine receptor 3 (ACKR3) lacks the ability to activate heterotrimeric G proteins and its activation leads to platelet inhibition and attenuates thrombus formation. In nucleated cells, heterodimerization of ACKR3 with CXCR4 regulates CXCL12-dependent signalling. The aim of our study was to investigate the formation of ACKR3/CXCR4 heterodimers in platelets and the subsequent consequences for platelet function.

Methods and results: Using a proximity ligation assay (PLA, Duolink®) to screen for CXCR4/ACKR3 heterodimerization inducing compounds, we found that ACKR3 agonism but not conventional platelet agonists or endogen ligands lead to heterodimer formation. To further characterize the formation of ACKR3/CXCR4 heterodimers, we studied the CXCL12-dependent platelet activation via CXCR4. Both, CXCL12-dependent platelet aggregation and collagen-dependent ex vivo thrombus formation were significantly downregulated by ACKR3 agonism. Moreover, platelet intracellular calcium and Akt signalling were increased by CXCL12 and again suppressed by ACKR3-specific agonists. Previously, CXCL12 was shown to decrease platelet cAMP levels via CXCR4. Treatment with a specific ACKR3 agonist counteracted this CXCL12/CXCR4-dependent cAMP decrease.

Conclusion: Our results reveal that the formation of platelet ACKR3/CXCR4 heterodimers is dependent on ACKR3 rather than CXCR4. Furthermore, ACKR3 agonism induced heterodimerization is associated with mitigating CXCL12/CXCR4-dependent platelet activation possibly by modulating CXCR4-dependent G protein signalling. Our results indicate possible ACKR3 agonist functions and reinforce the potential therapeutic applications of ACKR3 agonists.

Keywords: C‐X‐C motif chemokine ligand 12; C‐X‐C motif chemokine receptor type 4; atypical chemokine receptor 3; heterodimerization; platelet.

FIGURE 1

Heterodimers of ACKR3 and CXCR4 occur on platelets surface upon ACXCR3 agonism. (A) Scheme of proximity ligation assay Duolink® functioning. ACKR3 and CXCR4 are labelled with specific primary antibodies of different species. These are detected with the corresponding PLA probes with short reverse DNA tails, which are linked in the ligation step if the two receptors are in close proximity to each other. The signal is amplified by a polymerization reaction. Created with BioRender.com. (B) Representative images of PLA (magenta) and phalloidin (green) staining of untreated platelets or treated platelets with 5 μM ADP, 1 μg/mL CRP‐XL and 1 U/mL thrombin for 30 min at RT (upper panel) or with 1 μg/mL CXCL12, 1 μg/mL CXCL14, 1 μg/mL MIF (middle panel) or 100 μM ACKR3 agonist VUF11207 and C23 and control substance C46/DMSO (lower panel) for 15 min at RT. Scale bar = 5 μm. (C–E) PLA flow cytometry analysis of ACKR3‐CXCR4 interaction. (C) Platelets treated with 5 μM ADP, 1 μg/mL CRP‐XL and 1 U/mL thrombin for 30 min at RT compared to untreated control. Upper graph: Statistical analysis of ACKR3‐CXCR4 PLA positive platelets [%], Plotted: Arithmetic means ± SD of PLA positive platelets [%], n = 6, statistics: RM one‐way ANOVA (compared to untreated); n.s. not significant, *p < .05. Lower graph: Statistical analysis of mean fluorescence intensity [MFI] of ACKR3‐CXCR4 PLA positive platelets. Plotted: Arithmetic means ± SD of PLA positive platelets [MFI], n = 6, statistics: RM one‐way ANOVA (black arrow: Compared to untreated); n.s. not significant. (D) Platelets treated with 1 μg/mL CXCL12, 1 μg/mL CXCL14, 1 μg/mL MIF for 15 min at RT compared to untreated control. Upper graph: Statistical analysis of ACKR3‐CXCR4 PLA positive platelets [%], Plotted: Arithmetic means ± SD of PLA positive platelets [%], n = 6, statistics: RM one‐way ANOVA (black arrow: Compared to untreated); n.s. not significant. Lower graph: Statistical analysis of mean fluorescence intensity [MFI] of ACKR3‐CXCR4 PLA positive platelets, Plotted: Arithmetic means ± SD of PLA positive platelets [MFI], n = 6, statistics: RM one‐way ANOVA (black arrow: Compared to untreated); n.s. not significant. (E) Platelets treated with 100 μM ACKR3 agonist VUF11207 and C23 and control substance C46 /DMSO for 15 min at RT compared to untreated control. Upper graph: Statistical analysis of ACKR3‐CXCR4 PLA positive platelets [%], Plotted: Arithmetic means ± SD of PLA positive platelets [%], n ≥ 7, statistics: RM one‐way ANOVA; **p < .01. Lower graph: Statistical analysis of mean fluorescence intensity [MFI] of ACKR3‐CXCR4 PLA positive platelets. Plotted: Arithmetic means ± SD of PLA positive platelets [MFI], n ≥ 7, statistics: RM one‐way ANOVA; *p < .05, **p < .01, ***p < .001, ****p < .0001.

FIGURE 2

ACKR3 agonist‐induced heterodimerization of ACKR3/CXCR4 is associated with attenuation of CXCL12‐dependent platelet function. (A) Representative light transmission aggregation curves (left) and statistical analysis (right) of platelets treated with .1, .5, 1.0 and 5.0 μg/mL CXCL12. Plotted: Arithmetic means ± SD of platelet aggregation [%], n = 5, statistics: RM one‐way ANOVA, not indicated = n.s. not significant, **p < .01. (B) Representative light transmission aggregation curves (left) and statistical analysis (right) of platelets pretreated with 50 μM ACKR3 agonist (VUF11207 or C23) or 50 μM control substance C46 for 15 min at 37°C and activated with 1.0 μg/mL CXCL12. Plotted: Arithmetic means ± SD of platelet aggregation, n = 6, statistics: RM one‐way ANOVA, not indicated = n.s. not significant, *p < .05. (C) Statistical analysis of CXCL12‐dependent light transmission aggregometry in combination with threshold and high concentrations of ADP (left), collagen (middle), and TRAP (right). Plotted: Arithmetic means ± SD of platelet aggregation [%], n ≥ 3, statistics: RM one‐way ANOVA; not indicated = n.s. not significant, *p < .05. (D, E) Representative microscope images of DiOC₆ stained platelets (left, scale bar 100 μm) and statistical analysis (right) of ex vivo thrombus formation. D Washed platelets or E whole blood were pretreated with 100 μM ACKR3 agonist (VUF11207 or C23) or 100 μM control substance C46 for 15 min at RT and activated with 1.0 μg/mL CXCL12 for 10 min. Plotted: Arithmetic means ± SD of thrombus area fraction, n ≥ 4, statistics: Wilcoxon matched‐pairs signed rank test for not normally distributed data (black arrow: Compared to CXCL12 control); n.s. not significant, *p < .05, **p < .01.

FIGURE 3

CXCL12‐dependent platelet signalling is inhibited by ACKR3 agonism. (A, B) Intracellular calcium signalling of platelets using Fluo‐4 staining (30 min RT). (A) Representative microscope images (left, scale bar 5 μm) and statistical analysis (right) of washed platelets pretreated with 100 μM ACKR3 agonist C23 or 100 μM control substance C46 for 15 min at RT and activated with 1.0 μg/mL CXCL12. Plotted: Arithmetic means ± SD of Fluo‐4 intensity change, n = 5, ordinary one‐way ANOVA (black arrow: Compared to CXCL12 control); n.s. not significant, *p < ß.05, ***p < .001. (B) Statistical analysis of flow cytometry measurements of washed platelets pretreated with 100 μM ACKR3 agonist C23 or 100 μM control substance C46 for 15 min at RT and activated with 1.0 μg/mL CXCL12. Plotted: Arithmetic means ± SD of Fluo‐4 MFI signal, n ≥ 5, ordinary one‐way ANOVA (black arrow: Compared to CXCL12 control); n.s. not significant, *p < .05, ***p < .001. (C, D) Representative images of western blots detecting Akt phosphorylation at (C) T308 as well as (D) S473 of washed platelets pretreated with or without 100 μM ACKR3 agonists (VUF11207, C23) or control substance (C46) and subsequent platelet activation with 1 μg/mL CXCL12 and statistical analysis of the densitometric measurements of the AKT phosphorylation signals. Plotted: Arithmetic means ± SD of pAkt/Akt in percent, n = 8, ordinary one‐way ANOVA Dunnett's multiple comparisons test; not indicated = n.s. not significant, *p < .05, **p < .01. (E) cAMP levels of washed platelets pretreated with or without 100 μM ACKR3 agonist VUF11207 and C23 as well as control substance C46 for 15 min at RT and subsequent platelet treatment with 1 μM PGE₁ and 1 μg/mL CXCL12 for 10 min at 37°C. Plotted: Arithmetic means ± SD of cAMP concentration, n = 7 (CXCL12: N = 3; C23 and C46: N = 4), student's t‐test; not indicated = n.s. not significant, *p < .05, **p < .01.

FIGURE 4

CXCR4/ACKR3 heterodimer associated signalling in platelets. Schematic drawing of CXCR12 dependent‐signalling via CXCR4. ACKR3 agonism induced CXCR4/ACKR3 heterodimerization mitigates CXCL12/CXCR4‐dependent platelet adhesion, aggregation and thrombus formation by counteracting cAMP inhibition and attenuation of Akt and Ca²⁺ signalling. Created with BioRender.com.